> > should be harmless. are you running maybe
> > an old kernel?
>
> No, I downloaded the live-cd image about a few days ago.
>
> I also see the messages while running fossil only like:
> sdiahci: drive 1 in state ready
> sdiahci: drive 2 won't come up; in state new after 10 resets
> sdE2: wait
On Fri Feb 12 11:49:37 EST 2010, r...@swtch.com wrote:
> > acme doesn't always clear the right margin to compensate for
> > the different amount of right-margin slop due to different
> > character widths. in this case ">" is very wide and won't fit
> > in the space that "n" did. so the n is mista
>
> 2) Add support for more SATA/AHCI controlers; I have:
>
[...]
> 00:05.0 SATA controller [0106]: Silicon Integrated Systems [SiS] AHCI
> IDE Controller (0106) [1039:1185] (rev 03)
>
> is seems to be supported by the 9load and, or kernel, but I could not
> install neither the plan9 not 9atom o
> > 2) Add support for more SATA/AHCI controlers; I have:
>
> FWIW, GSoC projects that amount to ``add some drivers for my non-linux
> OS!'' have historically been unpopular, unfinished, and generally
> unloved projects (perhaps barring last year, when I didn't really pay
> attention). I recommen
On Thu Mar 4 17:15:19 EST 2010, david.eckha...@cs.cmu.edu wrote:
> > for example, there are a number of 10/100 chipsets that have
> > good documentation that's under 100 pages. complete with
> > (oh, my) a theory of operation.
>
> Such as?
and
On Thu Mar 4 20:28:36 EST 2010, ge...@plan9.bell-
> > * Gentoo's plan9port also sets DEPEND="x11-apps/xauth". Is xauth a
> > build or runtime dependency?
>
> I think it would be a runtime dependency, xauth pulls in all the X11
> libraries so you can X forward the GUIs from a headless machine. On the
> other hand, most of my binaries seem to be
> I think a broken table would make a better platform than a PC, and it
> seems to be getting worse.
this may be true. but can you name another platform?
arm is a cpu grab bag, not a platform.
> It certainly seems like there is a slight renewed interest in RISC
> machines.
alternatve interp
> erik quanstrom wrote:
> > perhaps the new cifs server would be better?
> >
>
> A pointer please.
>
> Already searched mailing list, (8) man pages, and the contrib packages
> but only found mention to aquarela as a cifs server.
it's in the distrib
> > it's in the distribution.
> >
> > cifs(4), /sys/src/cmd/cifs, /$objtype/bin/aux/cifs, /$objtype/bin/cifscmd
>
> Looks like the lookman index should be updated:
>
> term% lookman cifs
> man 8 aquarela # aquarela(8)
if you install even one man page that's not in the distribution,
you should
> OK. I dug a little deeper and found that rio uses x11-libs/libX11 and
> x11-libs/libXext as build dependencies. These I think are the correct ones to
> use and not xauth. This does raise the issue then if the gentoo ebuilds
> should conditionally use X and not build rio if you do not want rio
> I think you want to create some kind of redirected handle to the ctl file
> first, then start a new block in rc. I believe this is how network
> programming in rc can be accomplished in Inferno as well.
inferno doesn't have rc.
- erik
> > > I think you want to create some kind of redirected handle to the ctl file
> > > first, then start a new block in rc. I believe this is how network
> > > programming in rc can be accomplished in Inferno as well.
> >
> > inferno doesn't have rc.
>
> Sorry I meant the shell. I apologize to ev
> I've also started on updating an old per-application portage ebuild based on
> Anant's work back in 2007. If that is of any interest maybe I could get you
> to either collaborate or just do a little testing.
p9p is ment to work as a unit. i think it would
be better if you didn't.
- erik
> sys = penelopa
>dom=penelopa.karlov.mff.cuni.cz
>
> to /lib/ndb/local
>
> But that doesn't help. What's missing?
an ip address. also the spaces around the '='
mean that ndb will fail to see sys/penelopa
as a key/value pair.
since there is no ip address, dns is used.
and you have to u
> Btw. I saw the spaces in 'The Organization of Networks in Plan9' by
> Presotto & Winterbottom, part 4.1. on page 8, see 'sys = helix'. So
> there it is probably wrong...
it does look to be a typo. patch submitted.
- erik
> we've got fgb's wonderful program and I think we're crazy if we don't
> build on that.
>
> Or we're CADT.
haven't you heard? we're not allowed to do anything
for ourselves anymore. the mythical (and god like)
Library Writers do this for us. our job is to glue things
together and port:
http:
> Fgb's contrib sounds very good, I have not had occasion to try it but
> I presume it retains the scattered nature of the contrib directory.
[...]
> I'm looking at the problem from the single perspective of how it's
> done in NetBSD.
really? you haven't even tried it and your trying
to fit it in
> Thing is, the "port" hierarchy (hereto I used "NetBSD package system"
> for the same concept) provides both the hierarchical structure I
> believe is needed to minimise duplication and a description file to
> search for concepts rather than file names. So, yes, I agree with a
> portion of your s
> Now, it hands out DHCP addresses to my Windows clients, and, I can ping
> ns2.test.local by name, but, cannot ping the short name, ns2.
> This is because the client doesn't receive a "Connection-specific DNS suffix"
> from dhcpd. (You can see this in from "ipconfig /all").
> I thought that is
> Back to my original question: what currently uses the ipv6= tag? We
> need to update the man pages at least.
ndb/dnsquery. ndb/dnsdebug. it's inconsistent.
- erik
On Mon Mar 8 08:28:37 EST 2010, kokam...@hera.eonet.ne.jp wrote:
> > Any attempt to use venti too resulted in a failure sooner or later, as
> > was depicted in my posts. No hint why...
>
> Is your error messages indicate (1)cannot connected to venti server
> or (2)write error to some area in aren
>
> I thought the flashes themselves were doing wear-leveling these days in most
> products? That's not the case with sheevaplug? Or am I completely
> off-base?
>
there are a lot of embedded-space flashes that don't.
- erik
> my system is doing practically nothing. All 'stats' graphs are low
> with the exception of 'i'. I'm having like 1600 interrups per second
> (so I see a number instead of just the graph). Is this normal? How can
> I tell what those interrupts are?
if you are running 9atom, /dev/irqalloc 3rd colum
> Ok. I tried to cat /dev/irqalloc and by far (2 orders of magnitude)
> the highest number is
>
> 32 0 45425578 clock
>
> What is the meaning?
> Is this Ok?
yes. it should be fine. 9atom cranks up HZ to 1000.
this allows accurate sleeping delays down to 1ms.
this shouldn't be a problem, even
On Mon Mar 8 13:36:54 EST 2010, mistl...@yahoo.co.uk wrote:
> >> it supports 4gb of memory.
> > of non-ECC memory, so nice terminal, bad server
>
> Any recommendations of similar hardware which does support SECDED ECC?
the memory controller is integrated into the cpu
package (though not th
On Mon Mar 8 13:04:41 EST 2010, lu...@proxima.alt.za wrote:
> > You seem to insist on alien software, why is porting software from
> > lunix a prefered solution? Most of the solutions are a.) either to
> > problems we don't even have or b.) so awkward in interfacing you just
> > end up writing a
> Maybe Supermicro X8SIL
obviously not an atom, but
> Don't know how well this card is supported by Plan 9, but
i have tested it and everything works fine. of course
(repeating the op) the xeon processor is required to
get ecc, since the memory controller is part of the cpu.
and the core iX ver
On Mon Mar 8 16:35:12 EST 2010, st...@quintile.net wrote:
> perhaps of interest is seft [http://ww2.cs.mu.oz.au/~oldk/seft/]
> which works well for me. It is unusual in that it allows all the usual
> text searching tools (including AltaVista's long lamented () "near" operator,
> but does not use i
On Mon Mar 8 18:30:34 EST 2010, st...@quintile.net wrote:
> > why would we go for "slower than grep"? wouldn't it be simpler
> > put a new queryish interface on grep á la 9fans.net/archive?
> > what am i missing?
>
> It allows a subtly different set of query tools which make
> sense to less rege
On Tue Mar 9 00:27:59 EST 2010, lyn...@orthanc.ca wrote:
> > But there ought to be a sane
> > alternative and it should not be anywhere as complex.
>
> There is: it's called POSIX.
surely your joking, mr. nerenberg!
- erik
> However, there is no that call in /sys/src/boot/pc/sdiahci.c file.
> Is this intended ?
> If so, what it does mean?
i would say that's likely a bug, and it's not true in
my current driver. it's a shame that geoff hasn't
picked up the newer stuff.
my current versions are
/n/sources/contrib/quan
> i would say that's likely a bug, and it's not true in
> my current driver. it's a shame that geoff hasn't
> picked up the newer stuff.
i apologize for the tone of this.
i'm frustrated with ahci problems. i think
many of them have been fixed, but there
are good reasons for the lag. and many
a
>
> I have 1600x1200x16 at work on an nvidia card.
>
i have not been able to get 32-bit modes working
with vesa, but 16 bit modes do alright. i think there
may be some sort of (frame) buffer calculation that's
off. large images can crash one's terminal, too, even
in 16-bit mode.
previously i
> How do other operating systems detect the available RAM? Could Plan 9
> use the same method?
plan 9 uses the same methods for detecting ram
everyone else does. perhaps we trust bios too much,
but 8mb isn't impossible, either.
try
a *noe820scan=1
at the 9load menu prompt.
- erik
> believe it or not, I even do that on 9vx. For many things, esp. things
> in port, it's good enough for me. In fact I did a "curried pipe" in
> 9vx just to try some things out.
i find development to go very fast on plan 9.
now that /dev/reboot $kern is working for me, i
can restart the kernel in
> what is going to be in $1, $2... Is ghi the contents of $2 or not?
this is easy to test:
; echo abcdefghi | sed 's/(abc(def)*)(ghi)/\1/'
abcdef
; echo abcdefghi | sed 's/(abc(def)*)(ghi)/\2/'
def
; echo abcdefghi | sed 's/(abc(def)*)(ghi)/\3/'
ghi
- erik
> Ok, thats fine. But
> ;9 echo abcdefdffghi | 9 sed 's/(abc(d?f)*)(ghi)/\1/'
>
> this I don't understand...
> maybe it's too late for me...
you're re doesn't match at all. therefore no
substitutation is made. perhaps it would be
easier to see with this formulation:
9 echo abcdefdffghi | 9 sed
> > gcc's different % rounding
>
> hmm?
i'm sorry. totally wrong. i thought i remembered and i
didn't. it was a different problem entirely.
i found i had to make loads of changes like the following
/sys/src/9/ip/devip.c:47,53 -
/mnt/term/home/quanstro/hg2/vx32.old/src/9vx/a/ip/devip.c:47,53
> Thanks, Erik!
>
> At least I have found the iso image
> ftp://ftp.quanstro.net/other/plan9.iso.bz2 works quite correctly (even with
> venti!).
you're welcome.
- erik
> plumb start 9 cat /home/ruda/CALC/manu/$1_source | plumb -i -d edit -a
> action=showdata
>
> I am getting this message when it is used
>
> /home/ruda/CALC/systems/ni4ni4ni4_264_fcc001_2/rotxy/Cxy.eps
> /home/ruda/CALC/systems/ni4ni4ni4_264_fcc001_2/rotzx/Czx.eps
why are you executing a rule in
> This didn't work, the editor again tried to open the $data file, so I
> deduced 'plumb to edit' can't have additional parameters... (right?)
> ... and tried
>
> plumb start echo $data | plumb -i -d edit -a action=showdata
>
> which, however, does seemingly nothing... I don't understand why...
> Well. So now I just don't understand how the examples in
> Plumbing and Other Utilities by Rob Pike, page 11
> could work... Was there a rewrite? The example there is no longer
> valid... :( That cost me quite some time...
it is a very good question that should be answered.
for reference, the
i am now unable to copy more than about 10mb at a time
off the usb sd reader i've been using. usb/disk isn't just quitting,
the endpoint is dead. the console says only
usb/disk... usb/disk: : 062800
disk: reset: device is detached
it used to work, so maybe something new's come up.
oh, by the way, having the "debug on" option to the ctl file
was great. i love not rebooting. :-)
- erik
On Mon Mar 15 06:41:08 EDT 2010, n...@lsub.org wrote:
> That's weird, I didn't change anything recently.
>
> Do you know which change in sources broke your device?
>
no. in fact, now that i have a bit more experience with
this and reloaded the whole 2gb onto a usb hard drive, it
appears that it
it appears that sed won't handle an input line longer than 8k characters.
yet no diagnostic is printed. and this is not mentioned in the
man page.
example
for(i in `{seq 3000 3849}) x = ($x IMG_$i.JPG)
echo $x | sed 's/IMG_([0-9]+)\.JPG:\1:g'
no error is printed, yet the output
On Mon Mar 15 10:45:48 EDT 2010, n...@lsub.org wrote:
> usually, unplug, then plug.
>
> Otherwise you involve even usbd and the kernel on the process
> (you can do a port reset). The port reset code is in, but IIRC, only
> the keyboard is using it so far. It's a quite hard measure, so I'd prefer
>
On Mon Mar 15 13:30:52 EDT 2010, n...@lsub.org wrote:
> Ouch. Ok.
> I'll add a port reset upon babling errors for disk drives as soon
> as I have a bit of time for that. If it's urgent for you let me know.
>
thanks, now that you've given me a bit of a road map
of what the problem is and where an
> I'm having trouble installing on a clean disk with the latest iso (2010/3/12).
>
> I first removed all the partition info (using fdisk on a gparted liveCD) so
> that the disk was clear. I then booted Plan9, and when it asked to install a
> MBR, it returned an error that disk/fdisk failed. This
> fossil: diskWriteRaw failed: /dev/sdC0/fossil: score 0x: date ...
> part=superblock 0: i/o error
> fossil: diskWriteRaw failed: /dev/sdC0/fossil: score 0x0041: date ...
> part=label block 65: i/o error
>
> and it hung at 65% installed. I'll try a different partition.
those sc
On Mon Mar 15 17:12:06 EDT 2010, aim0s...@lavabit.com wrote:
> Just looked at source of wc
> (http://plan9.bell-labs.com/sources/plan9/sys/src/cmd/wc.c). UTF-8
> is hard-coded here. What is the reason? Nobody wants to rewrite it,
> it is optimization or it is impossible to rewrite it using runes fo
On Mon Mar 15 17:46:11 EDT 2010, aim0s...@lavabit.com wrote:
> On Mon, Mar 15, 2010 at 05:13:40PM -0400, erik quanstrom wrote:
> > perhaps you have misunderstood.
> >
> > inside programs, sometimes unicode text is represented as
> > runes. runes are not sent ove
> I'd like to create a directory /home and then bind it to /usr.
> (This is because I want some paths from linux that start with /home
> were valid paths in my plan 9, too.)
i just created my home directory in /usr on linux to solve this problem.
> However, I can't create a directory in my / ---
> The second problem arose when I repartitioned the disk with partdisk and it
> did not appear to have been updated when I continued with the install. I
> wonder if the partition map had not been updated with plan9's partprobe
> equivelent.
did you type 'w' then 'q' at the fdisk, the prep prompts
> And have noticed that an old image I use for qemu is going astray.
> Same kernel as it has been for quite some time, but the load is pegged
> at about 2500 at all times.
2.5, i assume. what are the 2-3 processes running?
- erik
> This image formerly ran in 256M, now requires 512M, because venti
> footprint is 140+211+211 ... wait, how does it ever fit in 512 anyway.
swap? this would answer two questions.
- erik
> I tend to disagree. If I'm running qemu it is because I want to
> simulate a whole-machine environment. If I don't need that simulation,
> I'll go back to 9vx.
seems that keeping up with qemu is at least as hard
as keeping up with real hardware.
- erik
> (1) It looks like to be able to read SATA CDROM, such that we can see
> directories, files at
> the top level of /n/cdrom. However, it fails to read the files under deeper
> levels,
> say such as /n/cdrom/sys/src/9/pc/pc. In short we cannot read cdrom
> correctly.
if you want to try the ne
On Wed Mar 17 11:17:05 EDT 2010, kokam...@hera.eonet.ne.jp wrote:
> I looked your new sources like a name of new-E820Atom Unfortunately
> it requires lots of changes to the structure of kernel device tree.
> Can't you rewrite them not dremand, say such as /sys/include/fits.h(?, sorry
> I don'
> I've been searching through the man pages and 9fans archive and I am
> unable to figure out how to correctly setup plan9 to read and write
> mail. I've added my mail servers in different places (i.e.
> /rc/bin/termrc, /lib/ndb/local, /mail/lib/rewrite.gateway), ran
> factotum, ran upas/fs but I'm
>
> my big question is "How do I get my employer to need plan9?"
>
by careful choise of employer? ☺
- erik
> > (2) vesa vga driver doen't work for say 1280x1024x16 or x 32 vgasize ( this
> > machine's
> > vga card is ATI Radeon HD 4350(1002/954f). the upper 1/4 part etc...
>
> This is still problem.
there may be some magic overflow crt register bits that
could be set incorrectly. i wonder about vga
> FWIW there is another weirdness of windows, the windows DHCP server doesn't
> communicate with the DNS server on windows, it expects the client to send an
> Inform packet to the DHCP server telling it of the clients chosen name.
i guess that's the great thing about standards — there are so
many
> Erik asked:
>
> > does anyone with a 2e licence know if plumber used to accept rules
> > like this; that is shell script fragments?
>
> It did not - because the plumber did not exist in 2e or earlier. Like
> a handful of other things, it was originally drafted for Inferno and
> then backp
awsome!
- erik
when viewing large images on a real plan 9 terminal,
i commonly see ilock deadlocks for example:
acid: src(0xf01b9432)
/sys/src/9/port/alloc.c:104
99 {
100Private *pv;
101
102pv = p->private;
103ilock(&pv->lk);
>104pv->lk.pc = getcaller
sorry, those were the pc and lockpc, respectively.
- erik
> I had this error on VMWare recently. The solution was to disable Venti and
> just use fossil. Venti is of questionable value when I'm already
> snapshotting my host's filesystem anyway.
won't help. his situation is dire. only 8mb of memory
seen.
- erik
> On 3/21/10, ron minnich wrote:
> > What's interesting to me about this is that
> > I can not imagine even attempting this on any other os or windowing
> > system. It was just too easy on Plan 9 however.
> >
> > ron
>
> so, are you basically saying that linux is a complex operating system,
> an
> So... how is the mail -c call in newuser supposed to work for normal
> (ie: not in group sys) users?
this is how the permissions were set up in 2005 on my machine:
; ls -ld /mail/box
d-rwxrwxrwx M 456741 upas upas 0 Jul 15 2009 /mail/box
- erik
> Have you tried Erik's alternative distribution:
> ftp://ftp.quanstro.net/other/plan9.iso.bz2
please don't use that! that's just the base image that
i change the kernel (and a few other things) from.
the correct path is
ftp://ftp.quanstro.net/other/9atom.iso.bz2
- erik
On Mon Mar 22 07:12:02 EDT 2010, jjackson...@gmail.com wrote:
> I'll let it sit for a while and see if it goes anywhere...
>
> Is the distribution you linked to anything like 9atom? Trying 9atom
> seems to be a common suggestion for these issues, so I gave it a shot
> and it freezes too, though th
On Mon Mar 22 11:59:21 EDT 2010, mirtchov...@gmail.com wrote:
> linus has won and we're just
> sulking :)
this all depends on your definition of "winning".
it appears your measure is user base. if this is
the definition, then windows has won.
but what i see is that there is increasing resistance
> Erik,
>
> That's really interesting...how do you keep up with all that hardware?
>
> I'll give it a try with a sata cdrom. Thanks!
it's the accumulated detrius of a few years' running with
various hardware. the original terminal i had at coraid
(which i still have) is a pentium iv with an ear
On Mon Mar 22 16:22:15 EDT 2010, n...@lsub.org wrote:
> It would help being able to "append" to a directory, i.e., being able to
> create new files but not to, say, remove, already created files.
i considered modifying the fileserver to do something similar
for nupas. perhaps reusing the 'a' bit.
> > russ had a specific objection to "append" directories. i don't
> > recall how the argument went.
> >
>
> Is it because you can bind in new stuff to a directory if you need it?
not really. it was related to the problems with /tmp directories
in linux.
consider a standard unix-style mailbox.
> Nope, I hardly ever tried nedmail. Others like Erik might know.
>
> jsyk, saving a message is dead easy from acme Mail, you just middle
> click on "Save imapfoldername" ;) (assuming the imap folder already
> exists)
i don't know about the current nedmail for p9p. i'd guess
that's a path proble
On Wed Mar 24 11:24:04 EDT 2010, n...@lsub.org wrote:
> About 4 * 40 here (students) plus
> our terminals and servers (4, 5 people, depends).
>
> On Wed, Mar 24, 2010 at 4:17 PM, Richard Miller <9f...@hamnavoe.com> wrote:
> >> I'm glad there's another person out there with 4 machines running plan
it appears that the problem EBo's been having has to do
with an inproperly set ccrp with amd sb7xx (and possibly
sb8xx southbridges). since the ccrp was not set properly,
sdiahci didn't recognize the ahci controller. this has been fixed.
many thanks to EBo for doing the hard work.
- erik
> Should there be something on the wiki/elsewhere where people can post
> the specs of computers they are using and what hardware works/doesn't
> work? The supported hardware list tends to be a bit vague; I'd like
> something where I could look at a list of components for an entire
> Plan 9 compati
> My Dell Inspiron 1000 mostly works. I haven't tested the PC Card slot. Audio
> most likely doesn't work, and the sis900 driver doesn't work and I haven't
> tested the other sis900 driver from here[1]. Everything else is fine.
could you define "doesn't work"? any errors?
- erik
> When I boot plan 9, the message `probing ether' is printed and then `auto neg
> something' is printed a second or two after that. I can find out what the
> actual
> message is later.
the output of /n/sources/contrib/quanstro/sosether
would be useful.
- erik
> From namespace(4), the man pages are supposed to be under /sys/man.
the old tex put the man pages (both of them) in /sys/man/1pub.
one could just as easily put them in /sys/man/1.
> What is the canonical way for added ("opt", "pkg" ?) stuff. Letting
> the user adapt his profile to bind the add
http://lwn.net/Articles/378219/
"[...] anything which combines tricky locking
and 30-line preprocessor macros is going to raise eyebrows.
But the core concept here is simple: [...]
oh, really?
- erik
> I'm inclined to this kind of stuff too, since it's easy to "rm -fr" if
> everything is simply in the same place.
replica should make such concerns moot.
- erik
> I've also noticed that the man page for plan9.ini(8), it says
> >On the SiS controllers, the Ethernet address is not
> >detected properly; specify it with an ea= attribute.
>
> Is ea= supposed to be set to the actual ip address (the man page says physical
> network adress)? If so that's probably
On Thu Mar 25 12:58:22 EDT 2010, rminn...@gmail.com wrote:
> It just keeps getting better:
>
>$ hugeadm --create-global-mounts
> $ hugeadm --pool-pages-max DEFAULT:8G
> $ hugeadm --set-recommended-min_free_kbytes
> $ hugeadm --set-recommended-shmmax
> $ hugeadm --pool-pages-min
> make sure you check out the final copy.c linked at the bottom of the page
don't follow this link. it is a trojan that will eat
into your brain and turn it into grey goo.
☺
- erik
> > > Is ea= supposed to be set to the actual ip address (the man page says
> > > physical
> > > network adress)? If so that's probably my problem.
> >
> > ea is supposed to be an ethernet address. ethernet
> > addresses have the format
> > ea=112233445566
> > where 1-6 are *lowercase* hex d
On Thu Mar 25 15:39:43 EDT 2010, lu...@proxima.alt.za wrote:
> > Bind your own directory on /bin with the creation flag set?
> >
> Nice idea, I like it.
on the other hand, doing this on a per-package basis
could quickly lead to chaos. to avoid this a whole set
of standards would likely be create
> 403 /* Output the newline. */
> 404
> 405 *bpout++ = '\n';
oddly, for such an obvious comment, it's not
exactly what the code does, and somewhat misleading.
that code just puts a newline in a buffer and increments
a pointer. outputting is elsewhere.
- erik
> I would like to see Plan 9 man support subdirs as rc does. For
> instance you can run ip/ping as a command, so why can't you look up ip/
> ping(1)? Man pages for add-ons would have their own subdirs under the /
> sys/man tree, and you would reference them with a syntax like package/
> pagena
> % echo 1.75e308+1.75e308 | hoc
> hoc 851: suicide: sys: trap: 19 (reserved) pc=0x3a75
trap 19 is SIMD floating point error. (sse or mmx.)
it's no longer reserved. it's quite curious that x87
floating point would generate such an exception in
plan 9 gives the following error (even with sse
> The Plan 9 approach today is either install everything in / (/386/
> bin, /sys/include, &c) or in your personal home dir and bind as
> needed. The later is irritating on multi-user systems, and the former
> can make maintenance a lot harder. Replica's -c and -s help, but it
> still r
> I see this as a little more complicated for two reasons. 1) administrating
> distributed systems with group supplied applications that are widely used but
> not part of the OS,
one of the key bits of plan 9 is that the fileserver has always been a
network fileserver. at coraid we have ~35 pla
> ok, what about libraries, are you going to add a new place for included
> files or have a script that sets the namespace for every library a
> program you want to compile needs?
>
> think about python for instances, it's needs zlib, bzip2, openssl and
> it couldn't depend even on more stuff (lik
> bind -a $home/bin/rc /bin
[...]
>
> (I use this for contrib packages as well, after getting burned a few
> times with contrib stuff breaking builds in /sys/src. Rather than
> use the package tool I copy the sources into $home/src and build as
> above. The extra work is minimal.)
how
> Well, I was expecting this question :-)
> But I don't actually have a good answer. It just felt wrong to let the
> program crash.
use notify(2) to do something with the signal.
- erik
> I'm doing the same here, same reasons. Works fine. And, yes, I had the
> same issues with contrib stuff breaking mk nuke all in /sys/src. I
> don't put any contrib src in /sys any more -- unless it is via bind.
why does one routinely do mk nuke?
- erik
> > how is this less work than fixing or removing
> > broken packages? that problem needs to be
> > dealt with one way or the other.
>
> Well, I've done both and this is less work. So I just measured the two
> quantities. :-)
adding contrib packages to the BUGGERED list
seems even easier. but s
1301 - 1400 of 6034 matches
Mail list logo
